Which audience analysis measurements principles and techniques are most effective, and why?

There are many methods to measuring audience opinion. The media uses probability sampling, quota sampling, survey evaluation, straw polls, interviews, diaries, meters, people meters, portable meters and Internet devices. Some of these tests are techniques and some are principles. I tend to think that the techniques are more useful than the principles, but that is just my opinion. However, one of the principles, probability sampling, is useful because it picks out a random group of a certain amount of people from that population which represent the the entirety of the group. This is useful because it allows the media to get the gist of a group without having to poll every single member. Another useful tool is interviews, if society is willing to participate in them. A company or station would call up a random household and ask if they would be willing to participate in a survey. Most often, this works better if there is a trade-off, such as a gimmick for the person who participating in the survey. But, when I take a survey, I am honest and reasonable, because I understand that they are just trying to improve their company. One of the other useful techniques is meters. These are useful because they track what channels a television set has on for most of the time. Overall, polling is probably the best method for the media to use to judge what the public enjoys most. All of these tools are somewhat accurate and can provide an ballpark figure of what mass audiences prefer.